GeneFlow: Translation of Single-cell Gene Expression to Histopathological Images via Rectified Flow

Spatial transcriptomics technologies can be used to align transcriptomes with histopathological morphology, presenting exciting new opportunities for biomolecular discovery. Using spatial transcriptomic gene expression and corresponding histology data, we construct a novel framework, GeneFlow, to map single- and multi-cell gene expression onto paired cellular images. By combining an attention-based RNA encoder with a conditional UNet guided by rectified flow, we generate high-resolution images with different staining methods (e. g. , H&E, DAPI) to highlight various cellular/ tissue structures. Rectified flow with high-order ODE solvers creates a continuous, bijective mapping between expression and image manifolds, addressing the many-to-one relationship inherent in this problem. Our method enables the generation of realistic cellular morphology features and spatially resolved intercellular interactions under genetic or chemical perturbations. This enables minimally invasive disease diagnosis by revealing dysregulated patterns in imaging phenotypes. Our rectified flow based method outperforms diffusion methods and baselines in all experiments. Code is available at https: //github. com/wangmengbo/GeneFlow.

Authors

• Mengbo Wang
• Shourya Verma
• Aditya Malusare
• Luopin Wang
• Yiyang Lu
• Vaneet Aggarwal Purdue University
• Mario Sola
• Ananth Grama
• Nadia Lanman

Keywords

No keywords are indexed for this paper.